Boosting cognition in older adults by means of foreign language learning

(1)

Research Master’s in Language & Cognition

Boosting cognition in older adults by means of

foreign language learning

A behavioral and electrophysiological investigation

Master’s Thesis

Floor Alicia van den Berg

s2691566

Supervisors:

Prof. dr. Merel Keijzer

Dr. Eleonora Rossi

Faculty of Arts

University of Groningen

The Netherlands

(2)

(3)

I hereby declare that this thesis is my original work, with due reference to the literature and acknowledgements of collaboration and assistance.

The present study was conducted under the supervision of Prof. dr. Merel Keijzer at the University of Groningen, and Dr. Eleonora Rossi at the University of Florida.

Floor van den Berg

Groningen, August 31, 2019

(4)

Acknowledgements

First and foremost I want to express my gratitude towards my supervisors Prof. dr. Merel Keijzer and Dr. Eleonora Rossi. I would like to thank both women for their inspiration, guidance, and support at every stage. I am grateful that I was given the chance to work on this intellectually-stimulating project, which has opened the gates for me towards exciting future opportunities. This research would not have been possible without the licenses provided by Rosetta Stone. Therefore, I also wish to thank Anita Bowles from Rosetta Stone for setting up the collaboration between Rosetta Stone and the University of Groningen, and Linda Hofman for her help in the process. I also sincerely wish to thank all participants for their perseverance and enthusiasm during their participation in this long and effortful study. I am also grateful to Ana Jimenez Tirado, whom recorded the Spanish stimuli that were used in this study. Thanks are also due to my fellow BALAB members for their helpful feedback, insights, and discussions on my work. A special mention goes out to Jelle, Vincent, Iris, Anne-Marie, and Peter, who kept me company during those long hours in the dark basement of the CNC and without whose help this exciting work would never have been possible. Last but definitely not least, I am thankful for my family and friends, who have kept me on my toes by asking the right questions and have always supported me in pursuing my ambitions.

(5)

Abstract

In this rapidly-aging world, dementia is predicted to quickly become the primary death cause among older adults. Therefore, preventative tools to combat age-related cognitive decline are attracting considerable interest. It is hypothesized that foreign language learning in older adulthood could be one of them, although much remains unknown about third-age language learning in general (Antoniou, et al., 2013). To this end, the aim of this work was to examine the effects of brief but intensive foreign language learning in a group of 64-78-year old healthy seniors on their cognitive functioning and well-being, as well as to investigate their foreign language learning progress and the relation between the two. Dutch older adults (n = 7) completed a battery of cognitive and linguistic tasks, and participated in three short tasks while their electroencephalogram was recorded, both before and after participation in a ten-day online Spanish course. While older adults showed significant improvements in accuracy in both Spanish production and comprehension, they did not improve in relation to general well-being or attentional control. Although we identified small possible vocabulary learning effects in ERP waveforms, no significant differences were observed. Lastly, working memory capacity predicted overall accuracy obtained in a Spanish vocabulary production task, and the number of languages learned across the lifespan was related to response latencies for cognates and non-cognates in that same task. Our study is the first to show that healthy older adults can quickly achieve vocabulary learning through an immersive online foreign language course. Additionally, it replicated the predictive effect of working memory on foreign language learning gains found in previous work. Although additional testing is required, our study contributes to the growing body of literature on third-age foreign language learning, and provides recommendations for research on how foreign language learning may boost cognition and consequently promote healthy aging.

Keywords: third-age language learning, healthy aging, cognitive functioning, individual differences, lexical processing, event-related potentials (ERPs)

(6)

List of Figures

Figure 1. Visual cues in the semantic categorization task... 32

Figure 2. 64-channel electrode array according to the International 10/20 system... 35

Figure 3. Accuracy scores and reaction times in the Spanish picture naming task ... 38

Figure 4. Cross-interaction between Session and Cognate Status for reaction times in the Spanish picture naming task ... 40

Figure 5. Number of correct items named in the Spanish verbal fluency task ... 40

Figure 6. Proportion of errors and reaction times per condition in the AX-CPT for the pre- and post-test ... 43

Figure 7. Behavioral Shift Indices in the pre- and post-test ... 45

Figure 8. Number of items named in the phonetic and semantic verbal fluency tasks for the pre- and post-test ... 46

Figure 9. Accuracy scores and reaction times in the semantic categorization task ... 49

Figure 10. ERPs of studied items in the pre- and post-test ... 51

Figure 11. ERPs of studied items and items not studied in the post-test ... 52

List of Tables

Table 1. Stimuli characteristics in the semantic categorization task. ... 31

Table 2. Conditions and their respective responses in the AX-CPT ... 33

(9)

2

List of Abbreviations

AD Alzheimer’s disease

AX-CPT AX-Continuous Performance Task CR Cognitive reserve

CRIq Cognitive Reserve Index questionnaire DTI Diffusion tensor imaging

EF Executive functioning EEG Electroencephalography

GM Gray matter

LLD Late-life depression

MCI Mild cognitive impairment MRI Magnetic resonance imaging PFC Prefrontal cortex

SCD Subjective cognitive decline

WAIS-IV Wechsler Adult Intelligence Scale-IV

WEMWBS Warwick-Edinburgh Mental Well-being Scale WM White matter

(10)

3

1. Introduction

The world population is aging rapidly - it is predicted that the proportion of older adults aged 60 and above is set to double from roughly 12% (2015) to 22% in 2050 (World Health

Organization, 2018). However, seniors are not necessarily aging more healthily. Next to declines in physical health, aging is associated with declines in cognitive and sometimes mental health, which could eventually evolve into memory-related aging diseases such as dementia or Alzheimer’s disease (AD) and increase the risk for late-life depression (LLD). In fact, it has been predicted that Alzheimer and other dementias will soon take over as primary population death cause number one, overtaking heart diseases as the current status quo (Kramarow & Tejada-Vera, 2019; Office for National Statistics, 2018). Considering the great challenge that these age-related diseases thus form, not only for the individual and their families but also for health organizations and governments worldwide, promoting healthy aging1_{is crucial in our current society. Therefore, it is imperative to investigate the potential} of behavioral preventative therapy to maintain cognitive health in the aging population, especially before cognitive decline progresses to clinical stages.

To maximize the benefits for cognition, interventions that call upon multiple cognitive domains have been argued to show most potential (Unverzagt, et al., 2013). One such intervention is especially interesting in that light: there is a promising hypothesis that learning a new language could serve as an anti-cognitive aging tool (Antoniou, Gunasekera, & Wong, 2013), because language learning involves an extensive brain network that overlaps with the network that is known to deteriorate with aging (Rodríguez-Fornells, Cunillera, Mestres-Missé, & Diego-Balaguer, 2009; Raz, 2000), and because language learning pertains to multiple cognitive domains as a highly complex new skill that competes with earlier stored languages (Pot, Keijzer, & de Bot, 2018). Until now, studies investigating changes in the brain as a result of foreign language learning have been mostly performed with college-aged

populations, and studies on foreign language learning in older adulthood have been scarce, despite Antoniou et al.’s promising hypothesis. However, late-life language learning in general is rapidly gaining in popularity and a research tradition is starting to emerge (e.g., Gabrys-Barker, 2018).

The few existing longitudinal studies on the cognitive effects of late-life foreign language learning have yielded mixed results (Klimova, 2018). Furthermore, there is very little scientific understanding of the behavioral and neural bases of the early stages of foreign language learning, especially in older adults. Finally, from an applied linguistics perspective, it is as yet unknown what the optimal teaching method for seniors is. It is of importance to

1_{Although ‘healthy aging’ is generally used to refer to physical health (Pachana, 2017), the present} paper uses the term to refer to physical as well as cognitive, mental, and social health.

(11)

4

investigate how methods can be tailored to older individuals and their preferences and characteristics to maximize the potential benefits of late-life foreign language learning on a linguistic (learning outcomes), cognitive, and social (well-being) level.

To this end, we present a pilot study, of which the goal is threefold. We aim to extend previous research on the potential benefits of late-life language learning by investigating 1) how online short-term intensive training in a foreign language may serve as a cognitive intervention program to improve cognitive functioning and general well-being in healthy older adults, 2) how baseline characteristics in cognition, affect, and linguistic abilities predict language learning outcomes, and 3) how foreign language abilities develop in older adults, which is examined by looking at behavioral and electrophysiological changes in auditory lexical processing after approximately 10 hours of language learning. This is a novel approach; the few published studies on foreign language geragogy have been exclusively behavioral in nature. Moreover, although a study operating under the same objectives has been done successfully with young adults (Eyer, 2018), the present study aims to extend existing research by targeting a healthy population of seniors.

2. Previous literature

In order to extend the existing work, it is necessary to first provide a theoretical overview of 1) structural, neurological, and cognitive changes in the aging brain and their consequences, 2) the effects of behavioral cognitive training on cognition and well-being in older adults, language learning and otherwise, 3) bilingualism and its effects on cognitive reserve, 4) the linguistic, cognitive, and neural consequences of foreign language learning, and 5) the characteristics of the third-age language learner, language learning outcomes, and their possible predictors.

2.1 Age-related cognitive decline

2.1.1 Functional and structural changes in relation to aging

Although not every person experiences cognitive aging to the same degree and at the same rate, the brain of every aging individual undergoes relatively predictable neuroanatomical and neurophysiological changes. Most notably, the brain decreases in size and weight as a function of aging, due to brain atrophy in both gray matter (GM), which supports

information processing, and white matter (WM), which regulates the transmission of neural signals between GM structures (Nagaratnam, Nagaratnam, & Cheuk, 2016). More

specifically, changes in neuronal structure associated with normal aging occur in the form of, for example, a decrease in the number and length of dendrites, demyelination, and synaptic loss (Murman, 2015). In other words, the connections between neurons in the brain become weaker, which has a wide effect on brain functioning and processing speed most notably. Some structures and networks in the brain are more susceptible to cognitive aging

(12)

5

than others (Cheng, Deng, Li, & Yan, 2015). Degeneration takes place especially in frontal regions such as the pre-frontal cortex (PFC), the area which is responsible for general executive functioning (EF), and in the temporal lobe, including the hippocampus, which is critically involved in (declarative) memory processes (Murman, 2015). As a consequence, on an overtly noticeable behavioral level, mainly EF and memory performance are affected by age-related declines in brain structure.

2.1.2 Cognitive changes in relation to aging

Processing speed and memory are most reported to decline as a result of age-related

neuronal deterioration in GM and WM (Ramírez-Gómez, 2016). These declines are generally manifested in an increased difficulty in the storage, retrieval, and retention of memories (Nagaratnam, et al., 2016). However, according to Nagaratnam et al., specific types of memory are affected differently. For example, the reduction of hippocampal GM volume results in deficits in the storage of short-term memories before they are consolidated into long-term memory. Both episodic memory (the learning and retention of new information) and prospective memory are affected, but procedural memory, which refers to long-term memories involved in learned skills, remains relatively intact.

In addition to memory processes, performance on tasks that require EF is also affected by cognitive decline (Murman, 2015). Older adults show declines in, for instance, attentional control (the ability to focus on specific information while ignoring irrelevant information), cognitive flexibility (the ability to flexibly transform, shape, and shift between information in response to changing demands; Miller & Cohen, 2001; Wilson, Nusbaum, Whitney, & Hinson, 2018), and working memory (Park, et al., 2002). Fluid intelligence, which refers to the ability to identify patterns and relationships, is also known to decline, whereas crystallized intelligence, encompassing breadth of knowledge, remains relatively intact (Craik & Bialystok, 2006), although the retrieval and control of this knowledge does become more difficult. With respect to language skills, speech perception, speech production, comprehension of meaning and vocabulary, grammar, and discourse are all negatively

affected by cognitive decline to some extent (Antoniou, et al., 2013; see Burke & Shafto, 2008 for an overview).

2.1.3 Variability in the consequences of cognitive decline

As mentioned before, although moderate cognitive decline is part of normal aging, older adults show high inter-individual variability in the onset of experiencing difficulties related to information processing and memory. It has been widely suggested that (life-long)

participation in mentally and physically-stimulating activities may result in a compensatory mechanism for these (neuropathological) declines in the brain, known as cognitive reserve (CR; Stern, 2012). CR is distinctly different from brain reserve: whereas ‘brain reserve’

(13)

6

specifically refers to the anatomical and physiological characteristics of the brain (e.g., volume, size, and number of neurons and synapses), CR pertains to the reliance on

alternative resources and more efficient processing strategies in the brain to manage brain pathology and to maintain normal cognitive function (Stern, 2012). Higher levels of CR may allow the aging brain to compensate for structural and functional decline, even when physical markers of dementia, such as neocortical plaques, are already present (Antoniou, et al., 2013).

In general, maintaining a healthy lifestyle (e.g., balanced diet, regular physical

activity, and healthy social environments) can positively contribute to healthy cognitive aging (Clare, et al., 2017). In addition, a higher educational level, occupational status, and

cognitively stimulating experiences may all uniquely contribute to cognition later in life (Opdebeeck, Martyr, & Clare, 2016). However, it is worth mentioning that individuals who initially display higher CR are more likely to experience a faster pace of cognitive decline once diagnosed with neurodegenerative diseases in comparison to those with lower CR (Hall, et al., 2007; Scarmeas, Albert, Manly, & Stern, 2006). This is attributed to the possibility that neuropathology will be more accumulated in older adults who were previously able to

cognitively compensate for this, resulting in a faster loss of function after the onset of clinical symptoms (Stern, 2009).

2.1.4 Risks for cognitive and mental health in older adulthood

CR can thus act as a compensatory mechanism for cognitive decline. Despite such a

mechanism in place, subjective memory problems in older adulthood may still occur, which may put individuals at risk for developing dementia, Alzheimer’s disease (AD), or other cognitive and memory-related disorders (Antoniou, et al., 2013). Indeed, subjective cognitive decline (SCD) may develop into mild cognitive impairment (MCI), which is best described as the transitive stage between normal aging and dementia (Nestor, et al., 2008). However, in approximately 30-55% of MCI cases, memory-related symptoms do not evolve into the disease-state of cognitive decline or may even reverse to normal (Sanford, 2017).

Additionally, subtle deficits in memory may precede the onset of AD by more than ten years, and decline may progress slowly and gradually (Antoniou, et al., 2013). It is therefore difficult to distinguish between the onset of memory-related disorders and normal cognitive aging. Consequently, it is possible that seniors who appear to have normal cognitive decline are actually gradually accumulating pathological decline and developing neurodegenerative diseases (Cheng, et al., 2015).

In addition to experiencing age-related cognitive decline, another (mental) health risk that older adults face is late-life depression. Late-life depression (LLD) is the term for depression diagnosed in older adults over 65 years old (McCall & Kintziger, 2013). In Europe, almost 30% of the elderly population suffer from depressive symptoms, which often relates to

(14)

7

the presence of chronic diseases, pain, mobility issues, and cognitive impairment

(Horackova, et al., 2019). Indeed, dementia patients are more likely to experience symptoms of depression, and individuals with SCD report more depressive symptoms than those

without it (Edmonds, Delano-Wood, Galasko, Salmon, & Bondi, 2014; Yates, Clare, Woods, & Cfas, 2017). Conversely, depressed seniors display cognitive deficits in EF more frequently than non-depressed older adults (Markides, 2007), most notably in the recruitment of cognitive flexibility (Johnco, Wuthrich, & Rapee, 2014). Deficits in cognitive flexibility in relation to depression are often linked to hypo-activity in the lateral and medial-PFC (Fitzgerald, et al., 2006; Remijnse, et al., 2013), leading to difficulties with emotion regulation and the tendency to ruminate (Genet, Malooly, & Siemer, 2013).

Apart from several physical factors, such as a reduction in sleep quality (Kurdziel, Mantua, & Spencer, 2017) and a general decrease in dopamine and serotonin levels (Smith, et al., 2009b; Troiano, et al., 2010), emotional and psycho-social factors may put older adults at risk for depression, too. The fact that a large number of older adults retire may lead to their experiencing reductions in their social networks and the absence of a life purpose, thus increasing the risk of loneliness and lower self-confidence (Ramírez-Gómez, 2016).

Considering the growing population of retired seniors in the world, the number of individuals suffering from depressive symptoms is prone to increase over time. Thus, it is crucial to implement ways to preserve mental health in older adulthood.

In addition to increasing older adults’ social support, social networks, and self-confidence (Wang, 2016), mental health could be promoted by reducing the hypo-activity present in the PFC, thus by increasing cognitive flexibility. Cognitive therapy aimed at increasing cognitive flexibility may not only influence the course of LLD, but could simultaneously battle cognitive decline and boost CR in the aging brain, consequently

decreasing the risk for developing neurodegenerative diseases. Studies comparing individuals with high and low CR have shown the benefits of CR for mental and cognitive health in older adulthood. For example, it has been shown that mood and rumination impact cognitive function in older adults with lower CR, but not in those with higher CR (Opdebeeck, Nelis, Quinn, & Clare, 2015). With respect to neurodegenerative diseases, older adults with high CR have been found to be almost 50% less likely to develop a form of dementia than seniors with lower CR (Landau, et al., 2012), again demonstrating the powerful protective effects that CR may have.

2.2 Cognitive training in older adulthood to promote healthy cognitive aging

Although it has been shown that engaging in cognitively stimulating activities throughout the lifespan can boost CR, an important question in the aging literature is whether the initiation of these activities later in life can still result in higher CR and consequently slow down, halt, or even prevent severe cognitive decline and depressive mood states. Any improvements as a

(15)

8

result of cognitive training would suggest that “the aging brain is very capable of neural reorganization” as a function of environmental demands (Park & Bischof, 2013, p. 112), also referred to as neural plasticity, similarly to younger brains.

This question has led to the development of (computerized) cognitive training programs aimed at older adults to promote optimal cognitive health. These programs are often confined to improving specific components of EF or memory. For example, Wolinsky et al. (2009) used the Advanced Cognitive Training for Independent and Vital Elderly (ACTIVE) program to assess the effect of training aimed at the improvement of memory, logical

reasoning, or processing speed on healthy seniors and older adults with suspected clinical depression. Only processing speed improved after training in the former group and resulted in participants being almost 40% less likely to develop clinical depression. However, the program did not significantly affect cognitive performance nor recovery patterns of suspected clinical depression in the pre-clinical group. Another study looked at the implementation of the same ACTIVE program, but this time in a group diagnosed with dementia (Unverzagt, et al., 2013). They found that the training did not result in a reduced incidence of dementia after a five-year follow-up. The authors proposed that this could be attributed to the short duration and intensity of the program (ten one-hour sessions over five to six weeks).

Other studies have found improvement of cognitive abilities that were specific to the training, and these effects lasted for months or even years post-training when regularly followed-up with booster or refresher sessions (as reviewed in Antoniou, et al., 2013; Murman, 2015). Incidentally, improvements may transfer to measures of memory and attention that were not targeted in the training program, and participants may even report noticeable changes in their cognitive functioning post-training (Smith, et al., 2009a). Thus, there is a possibility that some training programs may be effective in providing unique cognitive enhancements that are more difficult to gauge with other activities, but this may be highly dependent on the intensity and nature of the training. More specifically, a

combinatorial approach to cognitive training, targeting multiple cognitive domains simultaneously (e.g., executive functioning and memory processes), may have a higher chance of positively contributing to cognitive function later in life, thus being more beneficial in reducing risks for dementia and/or LLD (Unverzagt, et al., 2013).

Contrary to previous belief, seniors maintain a strong ability to learn new skills and retain new knowledge, provided that these skills are regularly practised (Park & Bischof, 2013). Because learning new skills involves advanced cognitive processing, new learning may make a significant contribution to CR (Fratiglioni & Wang, 2007). Although novel learning in older adulthood does become a more effortful process because of deficits in (working)

memory and processing speed (De Bot & Makoni, 2005), the hippocampus, which regulates memory, possesses a high capacity for neuroplasticity (Murman, 2015). Park and collegues

(16)

9

(2014) investigated the impact of learning new productive skills vis-à-vis practising receptive activities that did not require new (computer-mediated) learning on working memory, episodic memory, and reasoning in older adults. After a period of three months, courses tapping into productive activities such as digital photography and sewing had positively contributed to episodic memory, whereas receptive activities including listening to music or doing puzzles did not result in any improvements related to cognitive functioning.

Importantly, older adults in the photography intervention were required to master complex software and/or using the computer, and more fine-grained analyses comparing the

productive conditions revealed that only significant differences in episodic memory were reported in the photography condition and not in the sewing condition. All in all, this

evidence suggests that learning a new skill is related to cognitive stimulation in the sense that it boosts the aging brain, but only when this new skill is multifaceted and complex.

Taking all of this into account, it has been proposed that learning a new skill in older adulthood, i.e., a new language, can significantly contribute to healthy cognitive aging in a unique way because of its complexity, multidimensionality, and intensity (Antoniou, et al., 2013). This hypothesis results from the line of work that has shown that a life-time of using two or more languages, i.e., bilingualism, has broad effects on cognitive functioning and can protect against age-related cognitive decline.

2.3 The case of bilingualism

The past two decades has generated a vast amount of research examining the cognitive effects of bilingual experiences across the lifespan. This section synthesizes the most important findings in this field.

2.3.1 The bilingual mind

It has been widely suggested that the two languages in the bilingual brain, which are said to reside in a single language system, are continuously and simultaneously activated (Bialystok, 2017). Bilinguals must mentally control the activation of multiple languages in the brain by monitoring, selecting, and switching between these languages, depending on the context. Considering the unique omnipresence of language in life, and because (bilingual) language use requires a complex brain network, bilingual experiences may positively contribute to domain-general cognitive resources that are trained through their involvement in controlling multiple languages. Indeed, as a consequence of regulating two languages, young-adult bilinguals have been shown to outperform monolinguals on a range of non-linguistic cognitive measures, such as those tapping attention and conflict resolution (Costa, Hernández, & Sebastián-Gallés, 2008), cognitive flexibility (Kroll & Bialystok, 2013), and some components of working memory (Calvo, Ibáñez, & García, 2016; but see Paap, Johnson, & Sawi, 2016, who provide a different view on bilingual advantages). However, results are

(17)

10

often mixed for this age group, which may be explained by the possibility that young adults are at the peak of their cognitive abilities, consequently showing less variation in cognitive functioning than, for example, children or older adults, where cognitive functioning may be more compromised and where differences between monolinguals and bilinguals may emerge more clearly (Bialystok, Martin, & Viswanathan, 2005).

Despite this, neuroimaging studies have corroborated behavioral evidence by

showing that bilingual experiences shape the brain and result in enriched neural architecture. For example, young adult bilinguals have greater WM volume and integrity especially in the corpus callosum, leading to a more efficient exchange of information between hemispheres (Felton, et al., 2017; Pliatsikas, Moschopoulou, & Saddy, 2015). In addition, adult bilinguals display enhanced GM volume in bilateral frontal and parietal regions of the brain, which are involved in executive control, in comparison to monolinguals (Olulade, et al., 2016). These enhancements may be related to different or less brain activity in the PFC for bilinguals vis-à-vis monolinguals, suggesting the use of more efficient pathways by bilinguals to achieve successful information processing (Pliatsikas & Luk, 2016).

The advantages in EF observed for young-adult bilinguals in comparison to monolinguals have been found to increase with age, showing that bilingual advantages extend to older adulthood and may even be most pronounced at this life stage. It has been found that older-adult bilinguals display less steep age-related declines in cognitive

functioning than their monolingual peers (Bialystok, Craik, & Ryan, 2006; Bialystok, 2009) which may be corroborated by, for one, better maintained WM integrity (Luk, Bialystok, Craik, & Grady, 2011). These neural enhancements may transfer to behavior as well, as senior bilinguals have been found to outperform monolinguals not only on measures of episodic memory (Schroeder & Marian, 2012), but also on behavioral measures of cognitive control (Kousaie & Phillips, 2017), task switching, inhibition, and mental flexibility (Grant, Dennis, & Li, 2014).

2.3.2 Bilingualism and cognitive reserve

The fact that bilingual older adults show advantages in EF over monolinguals may be attributed to the possibility that bilingual experiences contribute to the build-up of CR, similarly to other cognitively-challenging life experiences. These advantages point to the possibility that bilingualism works as a protective mechanism against age-related cognitive decline, and this protective effect may even extend to the delay of neurodegenerative diseases such as AD (Bialystok, Craik, & Freedman, 2007): strikingly, retrospective studies show that some bilinguals show a delay in the onset of symptoms of dementia or AD by an average of 4-4.5 years, independent of education, sex, occupation, and immigration status, exhibiting the unique protective effects that bilingualism may provide (Alladi, et al., 2013; Bialystok, et al., 2007). Indeed, older adult bilinguals have been shown to be able to perform equally to

(18)

11

healthy monolingual adults on a range of cognitive tasks, even in the presence of GM atrophy as seen in AD patients (as reviewed in Grant, et al., 2014).

According to Grant and colleagues (2014), networks involved in both bilingual language control and executive control overlap in several brain areas, such as the PFC, inferior parietal lobule, anterior cingulate cortex, and the basal ganglia. The neural networks in these areas are also known to decline with aging. Thus, a lifetime of mentally juggling two or more languages, which recruits both language- and executive control processes, may have “far-reaching consequences for the mind and brain” (Grant, et al., p. 4) in older adulthood, presumably in the form of CR enhancement.

Most research in the field of bilingualism, its cognitive effects, and the consequences for CR has been performed with bilingual older populations already proficient in both of their languages in comparison to monolinguals. Despite the general trends that have been

observed in this kind of work, one of the major sources of mixed findings on the bilingual advantage is the fact that bilingual experiences are highly variable (de Bruin, 2019), which forms a great scientific challenge. For instance, next to age of acquisition or amount of

language use, the contexts in which languages are learned and subsequently used may greatly influence the consequences for the mind and brain (Green & Abutalebi, 2013; Takahesu Tabori, Mech, & Atagi, 2018). One way to control these inherently-variable bilingual

experiences is to introduce (monolingual) older adults to a bilingual experience later in life, i.e., foreign language learning, and collect data before and after the intervention. Such longitudinal work could simultaneously shed a light on the bilingual advantage debate, because it provides more insight into how the intensity or length of these bilingual experiences affect the timeline and nature of any cognitive benefits.

2.4 The consequences of short-term foreign language learning

Currently, most published longitudinal studies investigating foreign language development have examined college-aged populations. Previous research has argued that a substantial amount of input is necessary to see evidence of language learning, or that learning needs to take place before adolescence in order to reach high levels of proficiency due to a loss or reduction of brain plasticity in adulthood (Singleton, 2005). However, what threshold of language input is sufficiently substantial is unclear. Moreover, new evidence in relation to the age issue suggests that native-like brain activation patterns in response to a foreign language may emerge as a function of proficiency rather than age of acquisition (Green, 2003), and the brain has been found to show great malleability in function and neuroanatomy as a result of language learning far beyond puberty (Grant, et al., 2014). Nevertheless, there is still little insight into the early stages of language learning and when language learning begins to affect language processing and cognition within the individual language learner.

(19)

12

measures is electroencephalography (EEG), which measures electrical brain activity on the scalp. Due to its strong temporal resolution, the method is sensitive to the smallest changes in brain activity in response to external stimuli such as spoken or written language, also known as event-related potentials (ERPs), and is thus able to capture neural signatures that are specific to certain cognitive processes. An ERP component frequently associated with lexical-semantic processing is the N400, which consists of a negative deflection peaking at 350-550 ms post-stimulus onset, and is usually largest at centro-parietal sites on the scalp (Kutas & Federmeier, 2011).

The exact meaning of the N400 is still debated, but for the purpose of the present study we adopt the lexical-semantic integration account. This account posits that the

amplitude and latency of the N400 effect reflect the difficulty with which stimuli are accessed and retrieved from long-term memory (Lau, Phillips, & Poeppel, 2008), such that more negative or delayed effects generally indicate a higher difficulty in processing a stimulus. Several factors have been shown to affect lexical-semantic processing. For example, a delayed N400 is expected when leaners have acquired a foreign language at a later age and/or have a low proficiency in that language (Moreno, Rodríguez-Fornells, & Laine, 2008), because it would be more difficult to access lexical items from a low-proficiency language than from a high-proficiency language (such as the first language). Furthermore, cognate processing is facilitated in second language learners, such that cognates, which are lexical items that overlap in orthography and/or phonology between two languages, tend to elicit a smaller N400 amplitude than non-cognates (Midgley, Holcomb, & Grainger, 2011). Facilitation effects such as these have been shown to emerge very rapidly in the language learning process (Eyer, 2018). Lastly, overnight sleep may play a large role in the consolidation of new

vocabulary and lexical access (Bakker, Takashima, van Hell, Janzen, & McQueen, 2015a), such that recently-learned lexical items elicit fewer and less strongly pronounced negative deflections after a night’s sleep (although this lexicalization process may take longer with novel L2 vocabulary learning; see Keijzer, 2016).

A few studies have used (a combination of) behavioral and ERP methods to investigate foreign language processing in young adults before they reached considerable proficiency in that language. McLaughlin, Osterhout, and Kim (2004) investigated neural responses to real French words and pseudo-words in learners participating in an

introductory French course. Participants were presented with semantically-related or unrelated prime-target pairs, consisting either of a real word or pseudo-word. They were asked to judge the lexicality of each word pair (word/non-word) as part of a lexical decision task while their brain activity was recorded through EEG measurements. It was found that after only 14 hours of instruction in French, a larger N400 was elicited in response to pseudo-words in comparison to real pseudo-words, whereas behavioral performance remained at chance

(20)

13

level, showing that learners had become sensitive to French word structure after only a very brief period of instruction within the accompanying behavioral evidence. It was shown that this sensitivity increased over time. This suggests that ERPs may more accurately reflect (implicit) changes in language proficiency than behavioral data.

Language learning effects have been found after even shorter language learning interventions. Eyer (2018) examined the effects of a brief online Dutch course (approximately 10 hours of instruction) for native English learners who were either immersed in a Dutch environment or not. ERPs in response to lexical items, which were presented auditorily and had been taught during the course or not, were recorded while participants completed a semantic categorization task. Both groups were found to have acquired Dutch vocabulary, as evidenced not only from behavioral indices such as a decrease in RTs and increase in

accuracy scores, but also from a reduction in the ERP amplitude elicited in response to studied items following the language course vis-à-vis before learning took place. However, opposite ERP effects have also been observed in response to novel lexical items during the early stages of early foreign language learning (Yum, Midgley, Holcomb, & Grainger, 2014). The authors examined the acquisition of written Chinese lexical items in monolingual English speakers and recorded their EEG while the participants completed a go/no-go semantic categorization task, at four different points in time. They compared ERP waveforms for studied items across the four measurements, and related results to performance on a

translation recognition task, identifying a group of fast learners and a group of slow learners. For faster learners, they found that amplitudes became more negative at frontal-central electrodes as the training progressed, possibly reflecting more efficient mapping of the stimuli’s structural information to their meanings (Holcomb & Grainger, 2006).

Interestingly, slower learners exhibited a qualitatively different ERP pattern compared to the fast learners, such that studied items at later measurements elicited more positive waveforms at posterior scalp sites in comparison to earlier measurements. In other words, even when bilingual experiences are controlled in the sense of a foreign language course, individual differences emerge.

In addition to changes in language processing, neuroimaging studies have shown that foreign language learning can rapidly lead to changes in GM (Mårtensson, et al., 2012) and WM structure (Schlegel, Rudelson, & Tse, 2012) in language-related brain areas as a function of gained language proficiency. Mårtensson et al. found - by means of magnetic resonance imaging (MRI) - that cortical thickness had increased in areas related to the articulatory network after an intensive three-month interpreter training. In addition, interpreters showed an increase in hippocampal volume and increased cortical thickness in the left fronto-temporal cortex, which has been linked to rapid vocabulary acquisition (Davis & Gaskell, 2009; Ye, Mestres-Missé, Rodríguez-Fornells, & Münte, 2011). Furthermore,

(21)

14

language learning outcomes positively correlated with higher neuroplasticity in the

hippocampus and left superior temporal gyrus, suggesting that plasticity in these regions may be predictive of language learning success. Schlegel and colleagues investigated the

malleability of WM structure related to participation in a nine-month Chinese course by means of diffusion tensor imaging (DTI). Participants in the course progressively acquired efficiency in WM tracts in left-hemisphere language areas and their right-hemisphere

counterparts, as well as in frontal-lobe tracts, demonstrating the vital role that WM plasticity plays in foreign language acquisition.

Brief foreign language learning has also been shown to transfer to behavioral and electrophysiological indices of domain-general cognitive functioning. A short but intensive language course led to greater improvements in attention switching in young as well as older adults after 14 hours of instruction in comparison to participants partaking in non-language courses (Bak, Long, Vega-Mendoza, & Sorace, 2016). In addition, cognitive control has been found to become more efficient after a six-month introductory Spanish course, as evidenced from ERPs elicited in a go/no-go task reflecting an increase in attention (Sullivan, Janus, Moreno, Astheimer, & Bialystok, 2014), albeit in the absence of behavioral evidence. This is corroborated by several neuroimaging studies revealing that young-adult learners engage cognitive control and attention mechanisms especially in the early stages of second language learning rather than in later stages (Ellis, 2008; Grant, Fang, & Li, 2015). Another

component of executive functioning that likely plays a role in foreign language acquisition is cognitive flexibility (see Section 2.1.2). Cognitive flexibility may be required to solve the interference of one language on the other (most notably the influence of the first language on the new language) in order to achieve successful communication and language processing (Kroll, Bice, & Perrotti, 2015), and thus, foreign language learning may give cognitive flexibility a substantial boost.

In summary, language acquisition may occur very rapidly after only a short language learning trajectory. This may be manifested predominantly on a neural level especially in the early stages of learning, as the brain outpaces behavior here (McLaughlin, et al., 2004). Brief but intensive language learning, which acts as a mentally stimulating activity, may also increase behavioral performance on domain-general cognitive tasks, supported by neural adaptation in response to language learning. Although a general interest for the early stages of foreign language learning is emerging, much is currently unknown about early foreign language development in older adults, even though this is the age group in which language learning may be most beneficial. Do younger and older adults possess the same ability as young adults to acquire vocabulary rapidly, and how is their cognitive as well as mental health affected after only a few hours of foreign language exposure?

(22)

15

2.5 Foreign language learning in the third age

Three main findings can be gathered from the previous sections: 1) cognitive training can protect against (further) cognitive decline in older populations, 2) bilingualism has protective effects on cognitive health, and 3) cognitive changes can emerge rapidly in the very early stages of foreign language learning. These promising lines of research have recently been integrated to suggest foreign language learning as a cognitive therapy to prevent cognitive decline (Antoniou, et al., 2013). To reiterate, foreign language learning in later life may require a more extensive brain network than any other cognitively-stimulating activities, leading to improved connectivity between regions and an increase in brain volume and density in areas that decline with aging. More specifically, new language learning in older adulthood could enhance the brain mechanisms that support language processing, maintain the integrity of brain structures, and increase the number of potential neural circuits that compensate for cognitive decline (Antoniou, et al., 2013; Grant, et al., 2014). Importantly, not only could foreign language promote healthy aging in general, it can also provide an

interesting and meaningful pastime for (pre-)clinical older adults and third-agers alike, with the latter term denoting those older adults who are healthy, motivated, and live an active life after their retirement (Oxford, 2017).

2.5.1 The third-age language learner

As mentioned in Section 2.2, even the senior brain possesses malleability in response to environmental demands, perhaps even as a result of language learning. However, acquiring a second or foreign language in adulthood has long been deemed too challenging or even impossible (Birdsong, 1999). Indeed, past work gravitates towards an ‘optimal period’ for (second) language learning, which entails that the younger the age at which someone starts learning a language, the better they will be at acquiring that language (Paradis, 2009). This translated into society’s idea that starting to learn a new skill after adolescence is unlikely to be successful because of age-related declines in cognitive functioning, which in turn highly influences seniors’ own ideas about their ability to learn (Andrew, 2012). This may

consequently decrease their self-esteem and confidence in the language learning classroom (Ramírez-Gómez, 2016).

That is not to say that third-age language learning does not deviate from language learning at an earlier life stage. As a result of age-related cognitive declines in processing speed and memory, third-age learners may have more difficulty remembering new vocabulary, understanding complex instructions, and keeping relevant information in

working memory compared to younger adults (Ramírez-Gómez, 2016). Age-related deficits in working memory weaken the connections between novel information and information

already present in long-term memory, which makes it more challenging to transfer and recall this information. In addition, because consolidation of new knowledge into long-term

(23)

16

memory occurs during sleep, decreases in sleep quality may be paired with deficits in

vocabulary consolidation. Hippocampal-prefrontal connectivity is crucial in this respect, and thus age-related hippocampal and prefrontal deterioration may impair the consolidation process. However, the relationship between sleep and language acquisition in the third age has not yet been assessed on a structural basis (Kurdziel, et al., 2017).

Considering these characteristics of third-age language learners, in order to achieve a higher language learning performance plus associated cognitive and well-being

enhancements, it is imperative to make older adults aware of their own potential to learn in a more realistic and constructive manner such that goal setting may be improved (Ramírez-Gómez, 2016). Older adults have been shown to be generally more motivated and engaged in their language course than younger adults (Gómez Bedoya, 2008), but in order to maintain high motivation levels, courses should be tailored towards older adults’ needs and

preferences (Alvarado Cantero, 2008). Older adults vary in their instruction preferences, either because they are used to previous methods or highly disliked their previous language learning experiences (Assen & Busstra, 2018).

2.5.2 Third-age language learning outcomes

Despite the belief that it may be too difficult for older adults to acquire a new language, several recent studies have provided evidence for successful language learning by seniors. Most research in this respect has focused on the differences and similarities in language learning between young and older adults.

Lenet, Sanz, Lado, Howard, and Howard (2011) assessed the differences in acquisition of Latin morphology and syntax between college students and older learners (aged 66-81). Learners participated in a very brief computer-mediated language training (two grammar lessons of 30 minutes each), and were either enrolled in a training incorporating explicit feedback in the form of grammar rules or in a training where explicit feedback was absent. Language learning outcomes were assessed by means of aural interpretation, written interpretation, written grammaticality judgement, and written production tasks. The authors predicted that, because older adults tend to adopt different (and perhaps less effective) memorization strategies than their younger counterparts, seniors would benefit more from incidental and implicit instruction. Indeed, the young adult group benefitted more from explicit feedback overall, and in contrast, older adults significantly benefitted more from less explicit feedback, but only during the grammaticality judgement task. No differences in overall learning outcomes were found between the two age-groups, however, suggesting that older adults are capable of learning morpho-syntax as long as they are motivated. These result were replicated by Cox and Sanz (2015), who also found no differences in performance on Latin morpho-syntax between bilingual young and older adults after practising grammar rules. A follow-up study involving monolingual and late-bilingual older adults showed that all

(24)

17

participants showed significant gains on tasks involving interpretation, grammaticality judgement, and production, but bilingualism (in the sense of earlier foreign lanugage

learning experiences) provided an additional benefit regardless of treatment condition (Cox, 2017; also see Section 2.5.4). This suggests that novel language learning later in life is facilitated by earlier bilingual experiences.

Marcotte and Ansaldo (2014) aimed to investigate the neural mechanisms involved in vocabulary acquisition in French young (mean age: 22.7) and older adults (mean age: 70.2) by means of fMRI. Participants completed vocabulary training through a computerized Spanish word learning program, and underwent fMRI scans after 5 days of training (T1) and again when they had attained 100% accuracy on the program (T2), which on average was after 14 days of training for the young adults and 24.8 days of training for the older adults. Language gains were assessed by means of a picture naming task, during which participants named 80 Spanish nouns, comprising of 40 cognates and 40 non-cognates. Although accuracy was significantly lower for older adults at T1, all participants in both age groups eventually reached ceiling level in vocabulary performance, but older adults took

approximately twice as long as young adults to reach this level. In addition, non-cognate naming was significantly more difficult than naming cognates in both groups. Importantly, however, a larger cognate effect was found for older adults in comparison to young adults at T1, but the size of the effect became comparable for both groups at T2. Strikingly, the fMRI data revealed that older learners relied more on episodic memory and visual learning

pathways during lexical retrieval, rather than frontal cognitive control networks, which were primarily recruited by young learners. Thus, older adults may disengage prefrontal networks during lexical processing and show increased reliance on previous knowledge that is stored in long-term memory relative to young adults, which was corroborated by the stronger cognate effect for the former group at T1.

The studies discussed above all involved interventions focussing on a specific component of foreign language learning, such as morpho-syntax and vocabulary. Pfenniger and Polz (2018) extended this line of work to another language domain: they compared written production and comprehension (as measured by a C-test, odd-one-out task, and association task) of monolingual and bilingual seniors (aged 63-90) participating in a four-week intensive English language course (three lessons of two hours per four-week) teaching “the comprehension and use of familiar everyday expressions and very simple phrases to

communicate basic needs” (p. 4). Although no significant language learning gains were reported, both bilingual and monolingual participants improved their grammatical production and receptive vocabulary skills and made fewer errors as a function of the

training. Although there was a tendency for monolingual learners to show larger gains in the first few weeks of the language course, no advantages were observed for either group when

(25)

18

the course was completed, which does not converge with the advantages previously found for bilinguals (Cox, 2017).

Taken together, these studies show that language learning in the third age may be challenging, but far from impossible. In fact, considering that acquiring a foreign language in older adulthood is a cognitively-challenging activity and engages crucial brain regions that are known to deteriorate as a function of aging, such as the PFC and the hippocampus, third-age languthird-age learning may boost domain-general cognitive functioning in older adults. 2.5.3 Effects of third-age language learning on cognition and well-being

Since the publishing of the seminal Antoniou et al. (2013) paper, a few studies have aimed to relate third-age foreign language learning gains to improvements in behavioral indices of domain-general cognition and general well-being.

In addition to examining foreign language learning gains, Pfenniger and Polz (2018) assessed the effects of the English language course on inhibition, cognitive control, and socio-affective measures. Both monolingual and bilingual learners improved on inhibitory control and interference, as measured by the Stroop test. However, the training did not have a significant impact on measures of attention and cognitive control, as measured by the Alters-Konzentrations-Test (Gatterer, 1989), although many participants did report noticeable changes in their ability to concentrate. Regarding socio-affective factors, most learners indicated improvements in their general well-being after participation, reporting not only an increase in confidence and self-esteem, but also a sense of pride.

Cognitive effects have also been found after shorter language learning interventions targeted at seniors. For example, Assen and Busstra (2018) identified that a brief English language course (one hour per day for ten nearly-consecutive days) did not result in any improvements in self-reported general well-being, ascribed to the short duration of the intervention. Nevertheless, participants (aged 65-84) did display faster response latencies on incongruent trials and a fewer number of errors on the Flanker task, measuring inhibitory control, as a result of the language learning intervention. However, it is difficult to determine whether improvements reported in Assen and Busstra should be attributed to language learning or to practice effects (i.e., improvements due to having performed the task twice), considering the absence of an active or passive control group. Furthermore, as previously illustrated in Section 2.4, Bak and colleagues (2016) found that an intensive one-week Gaelic course (14 hours of instruction) significantly improved attention switching in both younger adults as well as older adults (age range 18-78). Importantly, Bak et al. identified a linear trend of improvement, such that participants in the language course improved the most, proportionally followed by those participating in other courses and participants who did not receive an intervention. Crucially, those participants who continued to practice their Gaelic for five hours or more per week were most likely to maintain their improved switching skills

(26)

19

nine months after the language training. Taken together, these results highly suggests that novel, cognitively-challenging activities may result in cognitive advantages (also see Park, et al., 2014), but that foreign language learning may indeed provide a unique cognitive boost. Nevertheless, foreign language learning programs targeted at older adults have not always been successful in improving cognition or well-being. In other words, this relatively new field is already characterized by mixed findings. Ramos, García, Antón, Casaponsa, and Duñabeitia (2017) examined task-switching performance in Spanish older adults before and after participation in a Basque language course during the course of eight months (three sessions per week for 330 minutes in total). Task-switching costs, as measured by a color-shape switching task, were not significantly reduced as a result of the language course. Similarly, in their pilot-feasibility study, Ware and colleagues (2017) implemented a four-month technology-based English language learning program paired with classroom

instruction for French older adults to evaluate effects on cognitive performance (assessed by the Montreal Cognitive Assessment) and subjective measures of loneliness and social

isolation (University of California Loneliness Assessment). Neither cognitive abilities nor measures of well-being significantly improved after the language course, suggesting that this computer-assisted learning approach was not effective in improving cognitive health or well-being in this sample of healthy seniors. Lastly, Berggren, Nilsson, Brehmer, Schmiedek, and Lövdén (2018) assessed verbal intelligence, spatial intelligence, working memory, item memory, and associative memory in a large sample of 160 healthy older adults participating in a randomized controlled study. The language intervention consisted of an 11-week Italian course (twice per week for 2.5 hours) and the active control group completed relaxation training (once per week for one hour). The authors found that none of the cognitive measures showed significant improvement over time, and concluded that foreign language learning is unlikely to lead to any improvements in cognitive functioning. However, it was relatively unclear why they opted to target these particular components of cognitive functioning and did not include indices of executive functioning such as cognitive flexibility or attention, which have been shown to be enhanced as a function of foreign language learning or bilingualism (see Section 2.3 and Section 2.4).

Antoniou and Wright (2017) describe several factors that may result in language learning gains and cognitive benefits for older adult learners, and in addition may explain differences in outcomes within the existing body of experimental work that has investigated the cognitive effects to emerge from third-age language learning. They suggest that a match or mismatch in language typology between a learner’s first language and the foreign language may result in different cognitive effects. In light of this, they propose two distinct roles for how language typology may affect cognitive benefits: 1) the processing complexity effect, i.e., the challenge of learning languages with distinct typologies poses a greater demand on

(27)

20

cognitive resources, or 2) the interference inhibition effect, i.e., learning languages with similar typologies leads to rapid language learning, generating cross-linguistic competition and consequently engaging executive control systems. Moreover, the authors hypothesize that the more languages compete in the mental lexicon, the greater the cognitive benefit may be. In other words, there may be an additional benefit of multilingualism over

monolingualism or even bilingualism when learning a new language (also see earlier in this section). The question then becomes what the bilingualism threshold would be: will

individuals who have had language training at one point in their lives show an effect, or is active bilingualism required? Furthermore, training intensity is mentioned as an important predictor of cognitive effects, as well as the context of subsequent language use (e.g., whether the environment allows for code-switching). Finally, the authors point out the risk of

comparing performance on different tasks that claim to measure the same EF components, as it is unknown whether these tasks indeed measure similar constructs. For example, it is difficult to compare the effect of language learning on switching performance in the participants from Bak et al. (2016) and Ramos et al. (2017), because the former used a switching task frequently employed in attention research, and the latter one a task tapping working memory.

In sum, it is currently unclear in which contexts and under what circumstance foreign language learning can be beneficial for cognitive and mental-health indices in older adult learners. This may be partly attributed to the fact that of the studies synthesized above, although intensive, few have tailored their language interventions to older adults’ learning needs.

2.5.4 Predicting language learning outcomes

Another, less prominent, line of research is concerned with the predictors of foreign language learning outcomes in older adults and tries to gauge their individual characteristics,

pertaining to cognitive, affective, and linguistic indices that may provide seniors with an advantage in acquiring a new language. Research outcomes can provide important

implications for the development of foreign language education programs targeted at third-agers to maximize the benefits for cognition and/or mental health.

Section 2.5.2 briefly tapped into the possible benefit of bilingualism on third-age language learning. Asking a similar question, Blumenfeld, Quinzon, Alsol, and Riera (2017) examined how previous linguistic experiences and cognitive skills predicted novel English language learning in older bilingual learners of various language backgrounds, who were immersed in an English-dominant environment. Cognitive testing before and after an English language course, which was specifically developed for the study and was taught one hour a week for 23 weeks, showed that performance on working memory (as measured by a digit span task) and orientation predicted gained knowledge of English. Importantly, the

(28)

21

authors also observed better learning in learners whose languages were dissimilar to English, thus showing that learning a language with a distinct language typology may contribute to larger language learning gains, as described by the processing complexity hypothesis (Antoniou & Wright, 2017). Another study also provided evidence for the role of working memory in third-age language learning by investigating nine Spanish older adults participating in a communicative English course, such that learners with higher working memory, independent of age, achieved higher language learning outcomes (Mackey & Sachs, 2012).

Along similar lines, a small-scale study investigated cognitive and linguistic

predictors of foreign language learning outcomes in older adults by means of implementing an English course (four lessons a day for a duration of three weeks) for ten monolingual German participants (Kliesch, Giroud, Pfenniger, & Meyer, 2017). Similarly to the studies discussed above, it was found that a better working memory, as measured by reading span and immediate verbal recall, were related to a steeper language learning curve. Importantly, the authors also identified fluency in the first language to be a significant predictor of language learning outcomes in their sample, suggesting that the preserved ability to

accurately retrieve and produce lexical items from the L1 lexicon resulted in larger language learning gains. The authors also collected qualitative data on the participants’ motivation to participate in the course and their motivation during the course. Besides personal

motivation, L2 progress and conscientiousness were mentioned as additional determinants in the participants’ willingness to participate. However, many older adults experienced fatigue, which the authors attributed to the intensity of the course.

Finally, we discuss the precursor study on which the design for the present study was based in more detail (Eyer, 2018). Besides showing that young adult learners of Dutch had consolidated Dutch vocabulary after approximately 8 hours of learning, this study revealed that fluency in the first language (similarly to Kliesch, et al., 2017) and the ability to balance proactive and reactive control (for a more elaborate discussion, see Section 4.5.2) were predictive of language learning outcomes. Bilingualism requires both the dynamic use of proactive and reactive control processes to monitor the language context, maintain the language selection, and to detect language switching cues (Green & Abutalebi, 2013), which may also be required during foreign language learning. However, Eyer did not observe any changes in behavioral indices of proactive and reactive control, probably due to the short duration of the language learning intervention. There is a possibility that changes in proactive and reactive control preferences are observed in older adults, considering their higher inter-individual variability in cognitive performance as compared to young adults, or that effects become apparent in ERPs, prior to behavior (similarly to Sullivan, et al., 2014).

(29)

22

wide range of individual characteristics, including but not limited to working memory capacity, fluency in the first language, socio-affective factors, and the learner’s language background.

3. Research questions and predictions

The present study aims to investigate the cognitive and mental effects, as well as the

predictors of short-term but intensive online foreign language learning in older adulthood by crucially combining behavioral and electrophysiological measures of cognitive functioning as well as auditory lexical processing. As the language learning intervention we implement a computerized short-term but intensive Spanish language course. As such, the following research questions and predictions form the basis of our study:

A. The influence of short-term foreign language learning on cognitive functioning and well-being in older adults:

1. Does short-term foreign language learning enhance executive functions (i.e., attentional control) on a behavioral level?

2. Does foreign language learning have an effect on self-reported measures of general well-being?

Considering the hypothesis put forward by Antoniou and colleagues (2013), and the fact that changes in executive functioning may emerge after only a short period of language learning in adulthood (e.g., Assen & Busstra, 2018; Bak, et al., 2016; Pfenniger & Polz, 2018), we expect that a short but intensive foreign language course affects cognitive functioning in older adults, particularly in the domain of attentional control, which incorporates processes related to cognitive flexibility (e.g., inhibition, planning, and working memory; Miyake, et al., 2000). The enhancement of cognitive flexibility in particular may lead to more efficient emotion regulation and less rigid thinking patterns, showing up overall as increased well-being levels. Therefore, we include self-reported mood and well-being as secondary variables in our study. Having said that, marked changes in self-reported mental health are not necessarily expected to emerge after a very brief period of language learning (cf. Assen & Busstra, 2018).

B. The behavioral and electrophysiological result of third-age short-term foreign language learning:

1. Does short-term language training through an individual, technology-based learning method lead to vocabulary learning?

2. What is the neural signature of auditory lexical processing as a result of short-term foreign language learning as measured using EEG?

Boosting cognition in older adults by means of foreign language learning

Research Master’s in Language & Cognition